KEY DATA: Starts: -14.4%; 1-Family: -2.4%; Permits: -5.4%; 1-Family: -0.8%/Claims: 280,000 (down 36,000)

IN A NUTSHELL:   “Home construction keeps bouncing around but with builder confidence soaring, it is likely the August slump will be followed by a September surge.”

WHAT IT MEANS:  Housing starts cratered in August but the Alfred E. Neuman in me holds strong: I am not worried.  July’s level was revised up to the highest rate in nearly seven years.  A 31.5% decline in buildings of five units or more was the major reason for the August drop and this component is extremely volatile.  For the first eight months of the year, starts are up by nearly nine percent, keeping up hopes that we could see another double-digit rise in construction activity.  I think that is likely for two reasons.  First, permits are still running above starts.  That points to an acceleration in construction.  Second, the National Association of Home Builders/Wells Fargo Housing Market Index surged in August to its highest level since November 2005.  Builders can get irrationally exuberant at times, but that is usually when construction activity is surging.  So look for a rebound when the September numbers come out.

With the Fed still focusing on labor, the sharp drop in the weekly jobless claims number was eye opening.  We are about as low as can be expected.  Don’t be surprised if this number soars soon.  The closing of three casinos in Atlantic City will likely mess up the data for a short time.   Also, the Philadelphia Fed’s Business Outlook Survey showed that activity grew at a somewhat slower pace in September.   Nevertheless, orders were strong, backlogs grew and hiring jumped.  Those details point to continued strength in the manufacturing sector.

MARKETS AND FED POLICY IMPLICATIONS: Housing continues to improve even if the gains are inconsistent.  Builders are a pretty confident bunch and that can only be because they are seeing activity pick up.  Thus, the fall off in construction activity should not be viewed as any sign of weakness.  With the labor market tightening, Janet Yellen may be repeating her point that if the data are stronger than expected, the Fed is prepared to act sooner than whatever the term “extended period” means.  Indeed, if housing starts do bounce back and manufacturing continues to grow strongly, that is precisely what the Fed will have to do.  Regardless, investors may be a bit confused by the inconsistencies of these numbers but that has never stopped them before.

A further thought on the Fed’s leaving in “extended period” in the statement.  Given her weird comments about the meaning of the phrase, that it was not calendar based but data based, I can only conclude that the Fed members would like to remove the words but only when they think the markets will not overreact.  They don’t want another 100 basis point gap up in rates.  I suspect that as soon as there are consistently robust job gains and the unemployment rate drops below 6%, which could happen by the end of the year, the phrase will be removed.  December is my guess.